V O L U M E 2 1 , NO. 9, S E P T E M B E R 1 9 4 9 Scott (3, 7’) gave a material of approximately the same molecular extinction coefficient as a sample obtained after two precipitations from alcohol by chilling, the latter contained less oxidation product as interpreted from the ratio of optical densities. That the product obtained by ammonia extraction is purer than any of the others is shown clearly by both a high ratio of optical densities and a high molecular extinction coefficient: 54,900 after one +nd 55,500 after two purifications. The material after the second purification was still impure, as shown by the fact that a carbon tetrachloride solution remained light yellow on extracting with 1 to 10 ammonia. Extraction of a carbon tetrachloride solution of the Eastman product was not very successful; however, the method may give good results. It is recommended that purification by ammonia extraction begin with alcohol-precipitated material. T o obtain a chloroform solution of exceptional purity, one can begin with an alcohol-precipitated sample and carry out several purifications by ammonia extraction from carbon tetrachloride. The di-pnaphthylthiocarbazole in the final ammonia solution can then be extracted with chloroform; the optical density a t 650 mp serves as a guide when one has a solution of serviceable strength. The molecular extinction coefficient (67,000) as obtained through the mercury (11) complex indicates that all the samples obtained by purification in this work and elsewhere ( 3 ) are still impure. Curve J of Figure 3 represents a chloroform solution in which 5 mg. per liter of di-8-naphthylthiocarbazone are completely converted to mercury (11)complex. KO absorption is shown a t 650 mp; therefore the impurities in the purified material cannot absorb a t this wave length unless they react with mercury (11) to form products that also do not absorb a t 650 mp. I n Table I1 the molecular extinction coefficients listed were obtained by the use of two stock solutions which differed widely in purity; samples 3 to 9 from a material having extinction coefficient of 55,500 and samples 10 to 16 from a material having a coefficient of 36,800. If impurities had reacted with mercury (11) the coefficients obtained through the use of the two solutions would have been
1139 different. That the change in optical density a t 650 mp is a measure of the amount of di-p-naphthylthiocarbazone which reacts with mercury (11)to form a colored complex having no absorptien a t this wave length, for concentrations employed, is shown by the data presented in Figure 3 and Table 11. The formula, Hg(DS)2, is assumed by analogy to the mercury (11) keto complex of dithizone. If the formulas are HgDN, Hg(DS)*,and Hg(DS)3 molecular extinction coefficients, employing the data from Table 11,would be 134,000,67,000, and 44,500, respectively. The purest sample prepared gave a coefficient of 55,500 and was shown still to contain some material other than di-p-naphthylthiocarbazone. Therefore the values of the molecular extinction coefficient given in Table 11, based on the formula Hg(DS)*, are the only reasonable ones under the conditions prevailing. Some values employing the zinc complex have been obtained, which, although they are somewhat more variable than those in Table 11, support the value of 6T,000 obtained through the mercury (11) complex. Work on mercury (11) and zinc as well as other complexes of di-p-naphthylthiocarbazone is continuing and the value of 67,000 for the molecular extinction coefficient in chloroform a t 650 mp is to be considered as tentative, LITERATURE CITED (1) Bamberger, E., Padova, R., and Ormerod, E., Ann., 446, 260-307 (1925). (2) Cholak, J., and Hubbard, D. hZ., IXD. ENG.CHEM.,ANAL.ED., 16, 333-6 (1944). (3) . . Ibid.. 18. 149-51 (1946). (4) Cholak, J., Hubbard, D. bl., and Burkey, R. E., >bid., 15, 754-9 1‘1943). --, (5) Fischer,-Emil, Ann., 190, 114 ( 1 8 i 8 ) ; 212, 316 (1882). (6) Hubbard, D. AT., ISD. ENG.CHEM.,ANAL.ED., 12,768-71 (1940). ( 7 ) Hubbard, D. M.,and Scott, E. IT7., J . Am. Chem. Soc., 65, 2390 (1943). \ - -
(8) Sullivan, M. L., S. C. L., doctor’s dissertation, Saint Louis
University, 1948. (9) Suprunovich, I. B., J. Gen. Chem. (U.S.S.R.), 8 , 839-43 (1938). RECEIVED November 19, 1948. Abstracted from t h e thesis presented by Vernon K. Kofron t o t h e faculty of t h e Graduate School of Saint Louis University in partial fulfillment of t h e requirements for t h e degree of master of science.
Microdetermination of Acetyl Groups Modi$cation of Elek and Harte’s Method R. B. BRADBURY Division of Zndustrial Chemistry, Councilfor Scientific and Zndustrial Research, Melbourne, Australia A n investigation has been made of the various stages in the microdetermination of acetyl, in which the acetyl group is hydrolyzed by aqueous p-toluenesulfonic acid and the acetic acid is distilled under reduced pressure and titrated iodometrically. The method used substitutes alkalimetric for iodometric titration, omits the unsatisfactory correction for sulfur dioxide, and directly avoids error due to carbon dioxide. It was successfully applied to acetylated alkaloids.
T
H E method of Friedrich, Rapoport, and Sternberg (2, 3 ) modified by Elek and Harte (1) involves hydrolysis by 2570 aqueous p-toluenesulfonic acid a t 100’ C., and distillation of the acetic acid a t 50 to 60 mm. into an ice-cold receiver containing 0.01 N iodine solution and potassium iodide. A correction for sulfur dioxide is applied by deducting the amount of 0.01 S thiosulfate used to titrate the iodine in the receiver from a blank using the same quantity of iodine solution. Iodine liberated by adding excess potassium iodate to the titrated distillate is assumed to be equivalent to the acetic acid together with twice the equivalent of the oxidized sulfurous acid. After standing for
20 minutes a t 35’ C. the solution is again titrated with thiosulfate The difference between the t6tal titer and double the equivalent of the iodine used in the first titration presumably gives the correct percentage of acetyl by the usual calculation. Investigation of this method showed that a true correction for acid decomposition products from p-toluenesulfonic acid cannot be obtained by reaction with iodine. Friedrich and Rapoport (Z), who first suggested this correction, later discarded i t (3). Suzuki (13) using a similar procedure made no such correction. Hurka and Lieb (6),after alkalimetric titration, attempted to determine sulfur dioxide by reaction with iodine,
1140
ANALYTICAL CHEMISTRY
although it may have been previously removed by boiling (10). I n the writer's experience, although ptoluenesulfonic acid alone gives insignificant amounts of acid decomposition products, yet bot,h SOa-- and SOa-- are formed in the presence of an acetylated compound. Sulfate, sometimes detected in the Kuhn-Roth method ( I O ) , also is disregarded in Elek and Harte's method. Iodometric titration of acetic acid is not free from objection. By using alkalimetric titrat,ion, the larger number of reagents required, necessitating their absolute purity, and also the time required for nearly quantitative liberation of iodine by acetic acid, are avoided. While Elek and Harte ( 1 ) alloxed only 20 minutes a t 35" C. for the liberation of iodine, Suzuki ( I S ) found that a t least 2 hours at room temperat.ure was neces-
7.5
7.3
1.1
5 6.9 6.7 3 AFTER ADDITION OF IML E K E 0.01 N Nop s z o ~OVER THE
6.5
EWIVALENT AMOUNT
6.3 6.1
0
5
sary. Kolthoff and Sandell ('7, 9) consider that acids having a dissociation constant greater than 1 X loeB can be titrated iodometrically-for acetic acid K, = 1.8 X Side reactions in the titration of iodine by thiosulfate are prevented when the pH is not greater than 6.5 and 5 for a 0.01 N and 0.001 X solution, respectively (8,9).
15
10
20
25 30 35 TIME, MINUTES
40
+ SapSZO8+ 5H20 = 2SaH SO4 + 8HI
-______
Table I.
51-
+
103-
+ 6H+ = 3H20 + 312
"Ifate
(1)
(2)
was followed potentiometrically a t 25" and 35" C. There was little difference in reaction rate for a 10" C. rise in temperature, the pH of the solution a t 20 minutes being between 7.0 and 7.1 in each case. However, when excess thiosulfate was added after 20 minutes' standing, the p H rose more sharply, reaching a higher value t,han in the first tR-0 experiments, and was still rising when the experiment was discontinued. Despite the conh u e d rise in pH (curves 1 and 2) it may be calculated that a t a pH of 7.3 attained a t 45 minutes, the reaction is approximately 99.775 complete, provided that. no side reactions occur. I n both iodometric and alkalimetric methods the effect of carbon dioxide is significant. Carbon dioxide has long been known to liberate iodine from an iodide-iodate solution (11), and this has been amply confirmed by the \Triter, using Elek and Harto's method. Either the carbon dioxide must be removed by bqiling t,he acid distillate (2, 6, 10) or, preferably, it must be strictly excluded from the apparatus. I n the former case the results achieved are obviously dependent on the concentrations of the acetic acid and carbon dioxide, and on the time of boiling. The ideal of removing all the carbon dioxide without loss of acetic acid is translated in practice into the compensatory removal of t,he bulk of the carbon dioxide together with a small amount of acetic acid. The loss of acetic acid for various boiling times had been studied by Friedrich et al. ( 2 ) and Hurka ( 4 ) ,who arrived at an optimum boiling time of 5 to 8 seconds. Although Friedrich et al. pointed out the significance of carbon dioxide in
60
55
Weight of BaSOde uivalent t o Sod--, gram Vol. of 0.01 N Ka8H equal to acidity derived from 0.5 gram of toluenesulfonio acid. ml. Sulfite Weight of BaSOdequivalent to SOa--, gram Vol. of 0.01 N NaOH equal to acidity derived from 0.5 gram of toluenesulfonic acid, ml. Total acidity, 0.01 N NaOH, ml. -
... .-
Effect of Decomposition of Toluenesulfonic Acid
Weight of p-toluenesulfonic acid
occurring in neutral solution, was favored by aorise in temperature, causing errors ranging from 1.84% a t 0 C. to 3.9% at 52" C. I n a comparison of iodometric, acidimetric, and potentiometric titration of acetic acid, Hurka (6) obtained values of 623, 640, and 658 micrograms per ml., respectively, on the same solution. For a titer of 5 ml. the difference between the iodometric value, which he assumed correct, and the potent,iometric, would amount to 0.28 ml. (5.670). The end point of t,he potentiometric titrat,ion was taken as 8.18. The change in the hydrogen ion concentration, according to the equation
50
Figure 1. Change in pH with Time
Pickering (18) found that the alternative reaction: 41,
45
Mannitol Nexaacetate, 30 .Mg. 4.887
S o Sample 4.883
0 001556 0: 14
0.001196 0.10
0.001173 0.10
0,000816 0.07
0.24
0.17
_ _ _ ~ ~ _ ~
__
Table 11. Effect of Carbon Dioxide Flask 1, Guard Flasks, Titer Titer M1. M1. 1. Distillate titrated without boiling 2. Distillate boiled for 5 seconds then titrated 3. Blank titrated immediately Athout boiling Blank boiled for 5 seconds,~thentitrated Blank stood for 3 hours. then titrated without boiling 6. Blank stood for 3 hours, then boiled for 5 seconds 7. Distillate heated just to incipient boiling, then titrated
!:
4.79 4.57
4.78 4.74 5.10
0.11 Si1
..
1.79
..
4.73
0.07
their iodornetric procedure, Elek and Harte found its effect to be negligible. The apparatus used by Elek aiid Harte was satisfactory, provided a constant pressure was maintained during distillation. Some acetic acid (0.05 to 0.19 ml., 0.01 N ) always remained undistilled because of the large surface area provided by the glass rod in the distillation flask. Variation in the concentration of p-toluenesulfonic acid from 12.5 to 50y0 or use of hydrolysis times greater than 3 hours had no effect. Sodium thiosulfate, 0.01 N , even when preserved with amyl alcohol, required daily checking. EXPERIMENTAL
Change in Hydrogen Ion Concentration during Liberation of Iodine. A solution containing 1.5 grams of potassium iodide, 5.00 ml. of 0.01 N acetic acid, and 45 ml. of boiled distilled water
V O L U M E 2 1 , NO. 9, S E P T E M B E R 1 9 4 9
1141
had a p H of 4.2 (glass electrode). Because the potassium iodide contained some alkali (1.5 grams requiring 0.19 ml. of 0.01 A' hydrochloric acid for neutralization) this value was somewhat higher than that of a standard 0.001 A' acetic acid solution (3.87). On addition of 2 ml. of 47, aqueous potassium iodate the p H rose ~-
'Table 111.
so,
Compound Acetanilide
2
Phenacetin
3
Mannitol hexaacetate
4
N , S'-Diace tylethvlenediamine 1,2-llethylenediosy-3-methoxy4-acetoxy-10methylacridone l-LIethoxy-2,3met hylenedioxy4-acetoxy-10methylacridone 1,2,3-Trimethoxy4-acetoxy-10methylacridone 1,4-Dimethoxy-2ethoxy-3-acetoxy-10-methylacridone 1-Methoxy-2ethoxv-3-acetoxy-4:hydroxy10-methylacridone 1.4-Diacetoxy-2ethoxy-3-methoxy-10-methylacridone 1-Methoxy-2-ethoxy-3,4-diacetoxy- 10-met hylacridone
5
6
7
8
9
10
11
12
13
14 15
16
17
18 19
20
21
Weight
CaHoON
l-Acetoxy-2,3-dimethoxy-4hydroxy-10methylacridone 1,2-Dimethoxy-3acetoxy-4hydroxy-10methylacridone 1,4-Diacetoxy-2,3dimethoxy-10met hylacridone 1-Acetoxy-2-rnethoxy-10-methylacridone-3,4quinone I-Acetoxy-2-niethoxy-10-methylacridone-3,4quinone azine 1,3-Diacetoxy-Pmethoxy-4hydroxy- 10methylacridone 1,3,4-Triacetoxy2-methoxy-10methylacridone 2-Acetoxy-3-methoxy-IO-methylacridone-l,4quinone 1,2-Diacetoxy-3methoxy-4hydroxy-10methylacridone l-RIethoxy-2,3diacetoxy-4hydroxy-10methslacridone -
_____
Anal? tical Results
Formula
1
-
4.290 5.897 7,209 '6,370 8.178 7.000 6.375 7,557 C1oH1sOzN 7.744 8.406 8.783 6.861 C'isHiaOir 5.299 7.315 4,252 5,230 CeHizO~X~ 4.598 4.489 C ~ E H I ~ O S S ( C H B C O9.046 ) 9,142
Vol. of 0.0100 s NaOH (Corr.)
Acetyl Found
3.23 4.46 5.34 4.78 6.09 520 4.70 4.33 4.41 4.81 4.92 3.93 7.37 10.18 5.87 7.21 6.36 6.22 2.67 2.71
32.4 32.5 31.8 32.3 32.0 31.95 31.70 24.6 24.5 24.6 24.1 24.7 59.9 59.85 59.4 59.3 59.45 59.6 12.7 12.7
CieHizOsK(CHsC0)
8,539 8.631
2.59 2.63
13.05 13.1
ClrHie01S(CHaC0)
8.281 9.254
2.40 2.68
12.4 12.4
CisHiaOaN(CHaCO)
9.139
2.55
12.0
CirHisOrN(CHaC0)
8,528 8,554
2.50 2.49
12.6 12.5
CiiHisOsN(CHaC0)z
8.508
4.36
22.0
C I ~ H I S O ~ X ( C H I C O )8.127 ~ 7,628 8.915 8,728 9.008 8,985
4.03 3.85 4.53 4.41 4.56 2.73
21.45 21.7 21.9 21.75 21.8 13.1
8,270 8.641
2.42 2.57
12.6 12.8
7.541 7,202
3.91 3.75
22.3 22.4
8.742 8.636 8,493
2.82 2.73 2.71
13.9 13.6 13.5
7 . 805 9.424
2.07 2.45
11.4 11.2
8.796 7.892 9.150
4.84 4.33 5.01
23.7 23.f323.00
9.333 8.366
6.75 6.02
31.0 30.9
9.224 9.399
2.91 2.96
13.6 13.5
8.536 6.645 8.826 5.913 6.878 6,554
4.73 3.69 4.82 3.26 3.71 3.55
23.5 23.9 23.5 23.7 23.2 23.3
Mithin a few seconds to 6.1. Thereafter the rise in p H with time w'&s followed a t 25' and 35' C. (Figure 1, curves 1 and 2). After 20 minutes a t 35' C., the conditions which Elek and Harte employed, the p H is between 7.0 and 7.1. Although the rate of removal by hydrogen ions is somewhat increased initially by a 10" C. rise in temperature, it becomes about equal in both __ cases a t 30 to 40 minuteb. After 7 to 10 minutes the reactiori is orily 99.1% complete (pH 6.8); virtual completion (99.9%) is not attained until after approximately Acetyl 2 hours (pH 7.7). Calcd. Where excess thiosulfate was added after %, 20 minutes' standing a t 35' C. (Figure 1, curve 31, 31.85 the p H rose to a higher value than in curves 1 and 2. D e c o m p o s i t i o n of p-Toluenesulfonic Acid. When 0.5 gram samples of microanalytical reagent 24 , ( I p-toluenesulfonic acid were submitted to distillation either directly or after 3 hours in a boiliug water bath, and the distillate was titrated, 0.15 to 0.16 ml. of 0.01-V sodium hydroxide was re59.4 quired for neutralization to phenolphthalein. A volume of boiled distilled water equal to 59.7 that in the receiver required 0.10 to 0.11 ml. of 0.01 3-sodium hydroxide. Thus, acid distillation 12.6 products from 0.5 gram of p-toluenesulfonic acid in the absence of sample required approximately 12.6 0.05 ml. of 0.011V sodium hydroxide. Carbon dioxide was completely excluded from the apparatus in these experiments. On distilling a 12.05 known amount of acetic acid in the presence of 0.5 gram of p-toluenesulfonic acid no extra 11.6 decomposition could be detected. In order to determine the extent of deconiposition in the presence of sample, the. coni12.05 bined distillates from 32 determinations after titration of the acetic acid were analyzed for sulfate and sulfite: sulfate as barium sulfate, 0.0031 21.56 gram, sulfite as barium sulfate, 0.0032 gram. Assuming a constant decomposition for every 21 3.5 sample, the total barium sulfate (0.0066 gram) is equivalent to 0.18 nil. of 0.01 S sodium hydroxide per determination. In order to confirm the fact that oxidation of 12 .i sulfite to sulfate wac' riot occurring during collrction of the residues, two experiments were carried 12.8 out. Approximately equal quantities of p-toluenesulfonic acid w t w heated with 2 ml. of distilled water for 3 hours in the presence and absence of 22.38 mannitol hexaacetate, and distilled, and the distillates were analyzed for sulfat,e and sulfite as 13.2 before. The precipitatioii of the sulfate as barium sulfate, however, was carried out in an oxygen-free 10.8 atmosphere. Tahle I lists the results obtained. These results show that somewhat mure sulfate than sulfite is produced by decomposition of p 23.2 toluenesulforiic acid. Effect of Carbon Dioxide. By distilling acetic acid in the presence of p-toluenesulfonic acid after 31.2 3 hours' heating at 100" C. and taking no special precautions to exclude carbon dioxide, variations 13.2 between titrations of up t o 0.5 ml. of 0.01 S sodium thiosulfate were obtained. On allom-ing acetic acid 23.2 to stand for varying periods of time in flasks covered with watch glasses, similar differences were observed. Several experiments were carried out 23.2 to check the boiling procedure recommended by some authors. Table I1 summarizes the results when 5.00 inl.
.
,
1142
ANALYTICAL CHEMISTRY
of 0.01 N acetic acid and p-toluenesulfonic acid were heated for 3 hours and distilled and the distillate was titrated. DISCUSSION
In view of the above investigations, Elek and Harte’s method was modified as follows: Attempts to determine sulfur dioxide n-ere omitted, because an estimate of the sulfur trioxide simultaneously produced could not be obtained. Gnder the conditions used by the writer, the amount of acid produced by the p-toluenesulfonic acid was approvimately the same as the holdup of acetic acid in the distillation flask, together with the amount of sodium hydroxide required to give the phenolphthalein end point in the volume of distillate obtained. Although these two errors compensated one another within narrow limits in this case, other workers would be wise to determine the holdup of acetic acid in their apparatus and the approximate acidity due to decomposition of their sample of p-toluenesulfonic acid, Iodometric titration of acetic acid was replaced by alkalimetric. Though it is conceded that carbon dioxide can be removed from the distillate by boiling, i t is considered more satisfactory to exclude i t from the apparatus by hydrolysis in a carbon dioxide-free atmosphere, and by using freshly boiled distilled water for the distillation and nashing of the condenser tube, etc. During distillation a t constant pressure the procedure was timed with a stop match t o avoid overheating when the contents of the flask m-ere nearly dry. Instead of using a 25V0 aqueous stock solution of p-toluenesulfonic acid, 0.5 gram of solid was weighed into the distillation Bask for each determination. Some of the results obtained by this modified procedure are indicated in Table 111.
ACKNOWLEDGMENT
The author is indebted to J. R. Price and W. D. Crow of these laboratories for permission to use hitherto unpublished data, and to the former for careful reading and helpful criticism of this paper. The work described in this paper was carried out as part of the research program of the Division of Industrial Chemistry of the Council for Scientific and Industrial Research, Australia. LITERATURE CITED
(1) Elek, A , and Harte, R. A,, ISD. EX. CHEM.,r i s . 4 ~ .E D . , 8 , 267 (1936). Friedrich, A., and Rapoport, S . , Biochem. Z., 251, 432 (1932). Friedrich, A,, and Sternberg, H., Ibid., 286, 20 (1936). Hurka, W., Mikrdchemie,30, 228 (1942). Ibid., 31, 5 (1943). Hurka, W., and Lieb, H., Ibid., 2 9 , 2 5 8 (1941). Kolthoff, I. M., Chem. Weekblad, 23, 260 (1926). Kolthoff, I . M., Pharm. Weekblad, 56, 572 (1919). Kolthoff, I. M,,and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” pp, 586-90, New York, Maernillan co., 1940. Kuhn. R., and Roth, H., Ber., 66, 1274 (1933).
Mellor, J. W., ”Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Vol. 11, p. 315, London, Longmans. Green and Co., 1937. Pickering, S.U.,J . Chem. Soc., 37, 128 (1880). Suruki, SI.,J . Biochem. ( J a p a n ) ,27, 367 (1938). RECEIVED April 19. 1948.
Technique in Semimicrodetermination of Uronic Acids GEORGE G. MAHER ‘Vorth Dakota Agriczdtural College and Experiment Station, Fargo, N. D .
HE procedure of Lefevre and Tollens ( I d ) has undergone Tseveral changes in the design of apparatus used and in the procedures follo~ved,especially in the macromethods ( 2 , 6-9, 13-15, 17, 18, 20, 21). A modification of the macrotechnique which nearly reaches the semimicroscope has been devised ( 1 0 ) and several modifications of 5 micromethod also exist ( 3 , 4). I n all these a general bulkiness in apparatus is objectionable and the manipulations are not so easy as they can be. The micromethod of Burkhardt et al. ( 3 ) requires an apparatus made by formidable glass blowing and involves a difficult titration technique. Its range of applicability is also questionable. 4 micromethod based on the absorption character of the product of reaction between uronic acid and napht,horesorcinol is available but unfortunately is applicable only to uronic acid solutions free of sugars whose reaction products absorb in the same region ( 1 1 ) . A method involving the gasometric measurement of carbon dioxide evolved ( 1 9 ) is very time-consuming and gives too divergent results on such simple uronic acid-rich materials as pectins ( 1 0 ) . Therefore the apparatus as described below has been devised and used on a semimicro scale. It is less elaborate than some of the preceding and is very simple to operate. APPARATUS A \ D R E AGEVTS
The apparatus is schematized in Figure 1. Absorption tower A , water condenser B , and scrubbing toiver C are fashioned from
ordinary Liebig air condensers of about 750-mm. length. Tower d is slightly pinched a t the lower end to hold the 0.6-em. (0.25inch) glass helices which pack i t to a height of about 250 mm. Tower C is similarly packed. Absorption flask D and reaction flask E are each of about 100-ml. capacity and preferably roundbottomed. Tube F is an ordinary drying tube with a piece of rubber tubing and a pinch clamp on the end. All glass tubing, except for those pieces on the safety bottle, is of 2-mm. bore. Rubber stoppers serve a t all points of attachment t o the towers and condenser. Fitting the stoppers, glass tubing, and towers together as shown in the drawing eliminates the possibility of any dead circulation spaces in the system. The glass tube running through the condenser terminates several inches below the tip of the condenser but at a point above the surface of t,he liquid in the reaction flask. Safety bottle S is of several liters’ capacity and is connected to a water pump. The bottle is fitted with a stopcock leak, L, and is connected to the rest of the system through a three-way, T-bore stopcock fitted wit,h an Ascarite drying tube. The only rubber tubing joint is at the junction between this T-tube and the outlet tube from the absorption tower. Heat can be applied through any desired means. A Glass-col heating mantle for a 250-ml. round-bottomed flask and a Type 200-C Variac voltage regulator were used by the author. This pcrmitted placing altheriometer bulb near the bottom of exterior of the reaction flask to facilitate reaction temperature control. The entire apparatus is mounted on a single ring stand. Other equipment includes a dispensing bottle, fitted with Ascarite and ;inhydrone tubes, for storing barium hydroxide solution; a wash bottle fitted wit,h a pressure bulb t,o which is attached an Ascarite tube: a delivery pipet of convenient size (the author used a 5-ml. automatic pipet with a side-arm filler tube and an overflow chamber); and a microburet of appropriate size.
